The journey begins with DevOps bottlenecks  - the backup of requests for resources. During his time as Head of Platform at BlueVine, a financial services technology company, Shaked Asakyo, CTO and Co-Founder of Kubuyia.ai, supported 200+ engineers with a small team of 5 DevOps and SREs. This situation led to DevOps bottlenecks - over 800 tickets a week in Slack and Jira and five sites to maintain in production. Shaked realized that the best way to relieve this painful exchange is to create a virtual Shaked - one that sits between developers and operators - that can work 24/7 in all time zones and has the full context to resolve the tickets.

Since 2017, Shaked’s first bot has evolved through the eras of machine learning and now Langchain and Large Language Models. 

This talk will discuss the use cases for AI in DevOps and also the learnings from using not just LLMs such as Llama 1 & 2, Claude, GPT 3,3.5 and 4, and Anthropic but also the usage of vector stores and classifiers to make AI more secure, predictable, and extendable.

This is a chance to see live examples of using AI to make DevOps more productive and less tedious.